Purchase this article with an account.
J. Lid, T. L. Morisbakk, P. O. Lundmark; Variables Affecting Short- and Long Term Repeatability of Macular Thickness Measurements Using Heidelberg Spectralis OCT. Invest. Ophthalmol. Vis. Sci. 2010;51(13):2289.
Download citation file:
© ARVO (1962-2015); The Authors (2016-present)
To assess short- and long term repeatability of macular thickness measurements using image aligned coherence tomography and to investigate variables that may affect repeatability.
38 normal healthy young volunteers underwent 3 repeated sessions with macular thickness measurements during a 2 months period (4 w intervals) by one experienced operator. Each session included 5 repeated measurements using radial- (90 and 180 deg) and raster scans in high resolution mode. The image alignment function (TruTrackTM) was activated after the 1st session. Average foveal thickness (FT) was calculated from 90 and 180 deg scans. Mean central (1 mm) retinal thickness (CT) and macular thickness (MT) within 8 predefined sectors were recorded from the thickness map in its default position. In addition, measurements were obtained with the thickness map centered manually at the thinnest location. The quality of each scan was graded according to sharpness and brightness of the IR image and to the resolution and quality of boarder definition of the thickness analyses of the OCT scan. The deviation of the OCT scan from a horizontal line was measured. Short- and long term repeatability was analyzed using Coefficient of Repeatability (CR), Coefficient of Variation and Intraclass Correlation Coefficient. In addition, the effects of scan quality and the deviation angle on the repeatability were analyzed.
Preliminary results (n = 9) indicate a general improvement of repeatability after the 1st session (CRFT = -19.1 µm; CRCT = -4.6 µm; average sectorial CRMT = -3.6 µm). However, because the repeatability of the 1st session was better with manual centering than in default position (CRCT = 3.3 and 7.3 resp.; average sectorial CRMT = 4.1 and 6.3 resp.), improvement of repeatability was less for CT and MT measurements when centered manually.
The results suggest that the repeatability of Heidelberg Spectralis improves with activation of image alignment and with manual centering of the thickness map when image alignment is inactive. Further analyses will explore the effect of scan quality and deviation angle.
This PDF is available to Subscribers Only